• Journal of Broadcast Engineering
• /
• v.24 no.6
• /
• pp.1113-1121
• /
• 2019

In this paper, we propose the face anti-spoofing method based on combination of luminance and chrominance with convolutional neural networks. The proposed method extracts luminance and chrominance features independently from live and fake faces by using stacked convolutional neural networks and auxiliary networks. Unlike previous methods, an attention module has been adopted to adaptively combine extracted features instead of simply concatenating them. In addition, we propose a new loss function, called the contrast loss, to learn the classifier more efficiently. Specifically, the contrast loss improves the discriminative power of the features by maximizing the distance of the inter-class features while minimizing that of the intra-class features. Experimental results demonstrate that our method achieves the significant improvement for face anti-spoofing compared to existing methods.

• Journal of the Korean Wood Science and Technology
• /
• v.47 no.3
• /
• pp.290-298
• /
• 2019

The sound absorption coefficient and transmission loss of several types of rice hull mats with varying apparent densities and thicknesses are estimated in this paper using the transfer function and matrix methods, respectively, to evaluate the possibility of using rice hull as an acoustic construction material. The mean sound absorption rates of 10-cm-thick rice hull mats with target densities of $0.10g/cm^3$, $0.12g/cm^3$, and $0.14g/cm^3$ were 0.91, 0.92, and 0.95, respectively, while those of the 1-cm-thick plywood attached to the back of the rice hulls were 0.90, 0.92, and 0.92, respectively. The means of the sound transmission loss of the 10 cm-thick rice hull mats with the target densities of $0.10g/cm^3$, $0.12g/cm^3$, and $0.14g/cm^3$ were 7.66 dB, 10.49 dB, and 14.14 dB, respectively, while those of the 1 cm-thick plywood attached to the back of the rice hulls were 33.34 dB, 36.72 dB, and 38.95 dB, respectively. In conclusion, a rice hull mat could be used as acoustic construction materials because of its high sound absorption coefficient and sound transmission loss.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.5
• /
• pp.1943-1948
• /
• 2014

Purpose: The prevalence of weight loss in esophageal carcinoma patients is high and associated with impairment of physical function, increased psychological distress and low quality of life. It is not known which factors may contribute to weight loss in patients with esophageal carcinoma during radiotherapy in China. The objective of this study was to identify the associated demographic and clinical factors influencing weight loss. Methods: We evaluated 159 esophageal carcinoma patients between August 2010 and August 2013 in a crosssectional, descriptive study. Patient characteristics, tumor and treatment details, psychological status, adverse effects, and dietary intake were evaluated at baseline and during radiotherapy. A multivariate logistic regression analyss was performed to identify the potential factors leading to weight loss. Results: 64 (40.3%) patients had weight loss ${\geq}5%$ during radiotherapy. According to logistic regression analysis, depression, esophagitis, and loss of appetite were adverse factors linked to weight loss. Dietary counseling, early stage disease and total energy intake ${\geq}1441.3$ (kcal/d) were protective factors. Conclusions It was found that dietary counseling, TNM stage, total energy intake, depression, esophagitis, and loss of appetite were the most important factors for weight loss. The results underline the importance of maintaining energy intake and providing dietary advice in EC patients during RT. At the same time, by identifying associated factors, medical staff can provide appropriate medical care to reduce weight loss. Further studies should determine the effect of these factors on weight loss and propose a predictive model.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.5 no.4
• /
• pp.57-64
• /
• 1985

An optimization model is presented that can be used in the determination of a loss rate function and conceptual runoff models using observed rainfall and runoff data. In order to estimate the lumped parameters and to control inputs of the model, the differential equations, linear for underground flow and non-linear for overland flow, are transformed into state equations. Parameters of a loss rate function and runoff model under stationary assumption can be determined by the following procedures: optimization technique, linear control and non-linear curve fitting theory using several multiperiod storms simultaneously or using individual multiperiod storms. An infiltration equation that includes rainful intensity is used to dtermine the effective rainfall for a given rain of varying. The optimization model is applied to storms in Hyong Song watershed of Wonju area. The results of the new model are compared with earlier one.

• ETRI Journal
• /
• v.42 no.1
• /
• pp.67-77
• /
• 2020

Image captioning has received significant interest in recent years, and notable results have been achieved. Most previous approaches have focused on generating visual descriptions from images, whereas a few approaches have exploited visual descriptions for image classification. This study demonstrates that a good performance can be achieved for both description generation and image classification through an end-to-end joint learning approach with a loss function, which encourages each task to reach a consensus. When given images and visual descriptions, the proposed model learns a multimodal intermediate embedding, which can represent both the textual and visual characteristics of an object. The performance can be improved for both tasks by sharing the multimodal embedding. Through a novel loss function based on class activation mapping, which localizes the discriminative image region of a model, we achieve a higher score when the captioning and classification model reaches a consensus on the key parts of the object. Using the proposed model, we established a substantially improved performance for each task on the UCSD Birds and Oxford Flowers datasets.

• The Journal of Korean Medicine Ophthalmology and Otolaryngology and Dermatology
• /
• v.21 no.2
• /
• pp.39-45
• /
• 2008

Objective : Skin barrier protects skin against harmful environment. Its function includes antimicrobial barrier as well as physical barrier. But there are few studies about the factors which improve skin barrier recovery after injury. The aim of this study is to investigate the factors which improve skin barrier recovery. Method : Nine hairless mice was anesthetized by inhalation and we tape stripped them. We used thermometer to know temperature one day ago, before anesthesia, before tape stripping, and after tape stripping. Vapometer was used to know transepidermal water loss before tape stripping, after tape stripping. And we used doppler flowmeter to measure blood flow before tape stripping, after tape stripping. Result : After analyzed data, we concluded that temperature was lower after anesthesia than before anesthesia and after tape stripping than before tape stripping. We could know that transepidermal water loss was lower after tape stripping than before tape stripping and more blood flowed after tape stripping. Conclusion : In our study, it was observed that temperature, transepidermal water loss, blood flow changed after tape stripping. But we thought lowered temperature was pathologic situation, more blood flow was to recover after injury. In traditional korean medicine, cold(寒) and imbalance of blood flow(血行) don't only make many skin diseases, but cold(寒) also obstructs blood flow. So we need to study how to warm cold(寒) and improve blood flow.

• ALGAE
• /
• v.35 no.3
• /
• pp.237-252
• /
• 2020

Temperate rocky reefs dominated by the giant kelp, Macrocystis pyrifera, support diverse assemblages of benthic macroalgae that provide a suite of ecosystem services, including high rates of primary production in aquatic ecosystems. These forests and the benthic macroalgae that inhabit them are facing both short-term losses and long-term declines throughout much of their range in the eastern Pacific Ocean. Here, we quantified patterns of benthic macroalgal biomass and irradiance on rocky reefs that had intact kelp forests and nearby reefs where the benthic macroalgae had been lost due to deforestation at three sites along the California, USA and Baja California, MEX coasts during the springs and summers of 2017 and 2018. We then modeled how the loss of macroalgae from these reefs impacted net benthic productivity using species-specific, mass-dependent rates of photosynthesis and respiration that we measured in the laboratory. Our results show that the macroalgal assemblages at these sites were dominated by a few species of stipitate kelps and fleshy red algae whose relative abundances were spatially and temporally variable, and which exhibited variable rates of photosynthesis and respiration. Together, our model estimates that the dominant macroalgae on these reefs contribute 15 to 4,300 mg C m^-2 d^-1 to net benthic primary production, and that this is driven primarily by a few dominant taxa that have large biomasses and high rates of photosynthesis and / or respiration. Consequently, we propose that the loss of these macroalgae results in the loss of an important contribution to primary production and overall ecosystem function.

• Journal of Broadcast Engineering
• /
• v.23 no.5
• /
• pp.614-621
• /
• 2018

This paper proposes a method of restoring corrupted depth image captured by depth camera through unsupervised learning using generative adversarial network (GAN). The proposed method generates restored face depth images using 3D morphable model convolutional neural network (3DMM CNN) with large-scale CelebFaces Attribute (CelebA) and FaceWarehouse dataset for training deep convolutional generative adversarial network (DCGAN). The generator and discriminator equip with Wasserstein distance for loss function by utilizing minimax game. Then the DCGAN restore the loss of captured facial depth images by performing another learning procedure using trained generator and new loss function.

• The Journal of the Acoustical Society of Korea
• /
• v.19 no.8
• /
• pp.12-21
• /
• 2000

In this paper, we propose a novel hearing aid algorithm for sensorinural hearing loss restoration with multi-channel(band) dynamic range compression and psychoacoustics. In this way, we can present a normal perception condition to the impaired listener. The proposed algorithm make loudness scaling function achieve proper loudness level, and analysis masking property for the signal will be perceived to impaired listener, and then, restore normal spectral contrast using SMR(signal-to-masking ratio) defined by distance between the level of each frequency and masking threshold.

• Proceedings of the IEEK Conference
• /
• 2002.07c
• /
• pp.1606-1609
• /
• 2002

MAC state models appeared with an effort to overcome technical demerits of CDMA in provisioning packet data service. In the scenario of sojourn and transition on MAC states, the design of state sojourn time is a critical issue for an efficient utilization of limited recource; a longer sojourn time leads to more resource being preserved for inactive stations, while more connection components should be recovered with a shorter sojourn time. Thus, the sojourn time at each MAC state must be optimized in consideration of these two conflicting arguments. In this paper, we first present a generic MAC state model. Secondly, based on the generic model, we reveal a relation of inactive period and the delay time of the last packet served in pre- ceding active period and specify a loss function reflect-ing two antinomic features that result from a change of state sojourn time. Using the proposed loss function, we construct a decision problem to find an optima3 rule for state sojourn times. Finally, we present a way of computing Bayes rule by use of the posterior distribution of inactivity duration for given observation on the delay time of last packet. Furthermore, Bayes rules are explicitly expressed for special arrival processes and investigated with respect to traffic load and loss parameters.